Many individuals, and especially adolescents, use one or more forms of social media. Social media companies, including Google, Facebook, Twitter, Microsoft, Pinterest, and Yahoo! do not provide outsiders with access to social media usage data, to protect user privacy and commercial interests. Even if one works for a large social media company, analyses of user activity across the boundaries of different sites are challenging or impossible. Investigating social media behavior and how it affects health behavior, including substance use and dependence, will benefit from studies that transcend social media boundaries. It also requires individual-level knowledge of participant behavior, health, and environmental context, which are difficult or impossible to obtain without individual-level clinical assessment. As such, it is currently unknown if and how social media exposure may relate to substance use and dependence, how this relationship evolves during adolescence, or the etiological sources of individual differences in this relationship. The present application is positioned to address this broad question through a combination of novel technology and standard in-person clinical assessment, all applied to a sample of adolescent twins from Colorado. The application will integrate in-person clinical, cognitive, and environmental assessment, online longitudinal surveys of the same, geospatial position logging, and monitoring of online and social media behavior. With these data, we will 1) monitor social media use and exposure for 400 adolescent twin pairs for a full year; 2) link aspects of social media behavior in these 400 twin pairs to known and validated in-person assessments of behavioral disinhibition, substance use and dependence, and environmental risk for substance use; and 3) evaluate the extent to which social media, GPS, and online monitoring can provide novel, accurate, and highly efficient assessments of substance use/dependence, behavioral disinhibition, and environmental risk factors for substance use/dependence. After validating the online data collection measures, we will test for longitudinal associations between indicators of social media use (e.g., duration of use, posting/reading/searching for drug- related content) and substance use, utilizing the twin design to parse the genetic and environmental contributions to these associations.